169 



The results now obtained for different dome thicknesses 

 are shown by curve B, fig. iv., which it will be seen can again 

 be represented approximately by the expression 500 + 



430e-r25d. 



It seems reasonable to suppose that besides the constant 

 normal leak in the ionization chamber, secondary y radia- 

 tion from the sides of the conical hole can enter the chamber, 

 and that this radiation may for the purposes of correction 

 to which it is to be applied later on, be taken as equivalent 

 to a homogeneous radiation for which A = 1*25. 



If now a Pb plug, 4 cm. deep, be placed in the coni- 

 cal hole in the position shown by a, fig. i., so as to intercept 

 the main stream of y radiation, but still leave the sides of 

 the cone exposed to the ionization chamber, and a set of ex- 

 periments similar to the above be performed, the results 

 shown in curve C, fig. iv., are obtained. These again are 

 approximately represented by 500 -}- 270e-i*25d- 



It is not surprising that the constant normal leak has 

 not been reduced to any appreciable extent, for it will be 



